| Time |
Nick |
Message |
| 00:25 |
|
Lesha_Vel joined #minetest |
| 00:37 |
|
sys4 joined #minetest |
| 00:59 |
|
sys4_ joined #minetest |
| 01:31 |
|
sys4 joined #minetest |
| 02:09 |
|
est31 joined #minetest |
| 02:10 |
|
fling joined #minetest |
| 02:30 |
|
est31 joined #minetest |
| 03:03 |
|
Alias joined #minetest |
| 03:06 |
|
est31 joined #minetest |
| 03:14 |
|
sys4 joined #minetest |
| 03:45 |
|
fling joined #minetest |
| 04:00 |
|
MTDiscord joined #minetest |
| 04:07 |
|
Verticen joined #minetest |
| 04:18 |
|
sys4 joined #minetest |
| 04:19 |
|
fling joined #minetest |
| 05:53 |
|
lemonzest joined #minetest |
| 06:01 |
|
calcul0n joined #minetest |
| 06:11 |
|
orwell96 joined #minetest |
| 06:22 |
|
wallabra_ joined #minetest |
| 06:36 |
|
sys4 joined #minetest |
| 06:52 |
|
definitelya joined #minetest |
| 07:15 |
|
est31 joined #minetest |
| 07:25 |
Izaya |
speaking of building, on debian stable, you need to use irrlichtmt 1.9.0mt4 to build 5.5.1 |
| 07:36 |
|
cranezhou joined #minetest |
| 08:20 |
|
erle joined #minetest |
| 08:58 |
|
appguru joined #minetest |
| 09:01 |
|
specing joined #minetest |
| 09:28 |
|
___nick___ joined #minetest |
| 10:10 |
|
Fixer joined #minetest |
| 10:21 |
|
Talkless joined #minetest |
| 10:54 |
|
x_ joined #minetest |
| 11:43 |
|
lemonzest joined #minetest |
| 12:04 |
|
erle joined #minetest |
| 13:36 |
|
vampirefrog joined #minetest |
| 14:10 |
|
z812 joined #minetest |
| 14:44 |
|
appguru joined #minetest |
| 14:59 |
|
perrier joined #minetest |
| 15:40 |
|
natewrench joined #minetest |
| 15:49 |
|
sys4 joined #minetest |
| 15:51 |
natewrench |
I think Mojangs stance on NFTs was virtue signaling |
| 15:55 |
rubenwardy |
? |
| 16:07 |
erle |
natewrench explain? |
| 16:13 |
natewrench |
they didnt have to elaborate about banning nfts i mean its there in their terms & EULA |
| 16:13 |
natewrench |
they got heat from the chat reporting so they did this |
| 16:14 |
natewrench |
the same day 1.19.10 is released |
| 16:15 |
erle |
come on, it's no surprise corporations lie to you |
| 16:16 |
erle |
just paint the logo in rainbow colors and you are good :P |
| 16:16 |
|
___nick___ joined #minetest |
| 16:16 |
erle |
(i wish i was joking) |
| 16:16 |
natewrench |
yea to think we own the worlds we create is bogus |
| 16:17 |
erle |
obviously, the end game is that disney owns it |
| 16:18 |
natewrench |
wait disney doesnt own minecraft? does it? |
| 16:21 |
|
KINGOWAIN_YT joined #minetest |
| 16:22 |
definitelya |
natewrench: Correct, Microsoft does. |
| 16:23 |
natewrench |
through their subsidiary mojan |
| 16:23 |
rubenwardy |
"all your world belong to us" |
| 16:24 |
natewrench |
yea you cant access you world without our client |
| 16:24 |
definitelya |
T_T |
| 16:28 |
definitelya |
natewrench: Wait by "our client", do you mean the official Minecraft one? |
| 16:28 |
natewrench |
yea |
| 16:29 |
definitelya |
There is also MultiMC, but it's still subect to Microsoft shenanigans. |
| 16:29 |
natewrench |
true |
| 16:29 |
natewrench |
optifine |
| 16:29 |
definitelya |
*subject |
| 16:30 |
MTDiscord |
<ROllerozxa> multimc is inching more and more into becoming proprietary |
| 16:30 |
natewrench |
even if its under the apache license it relies on MC |
| 16:30 |
MTDiscord |
<ROllerozxa> or basically, a code oss - vs code situation |
| 16:30 |
definitelya |
Ah, oh well I don't even have an account anymore. |
| 16:32 |
definitelya |
I just play on Minetest, it's free software and has cake and cookies, too. lol |
| 16:34 |
|
luiz joined #minetest |
| 16:35 |
|
M56[m] joined #minetest |
| 16:36 |
luiz |
Hi. I was trying basic_signs mod and I have found a error with craft some sign it requires steel:sheet_metal. I already install a mod called Steel |
| 16:36 |
luiz |
I have homedecor modpack too and signs_lib too |
| 16:48 |
|
fluxionary joined #minetest |
| 17:20 |
|
sys4 joined #minetest |
| 17:37 |
|
Yad joined #minetest |
| 18:09 |
|
WhiskeyPotato joined #minetest |
| 18:15 |
|
WhiskeyPotato joined #minetest |
| 18:29 |
WhiskeyPotato |
Hi everyone. So im having an issue with my minetest based server where the server freezes and the log is filled with: |
| 18:29 |
WhiskeyPotato |
2022-07-23 13:35:11: ERROR[Server]: Server::ActiveObjectMgr::addActiveObjectRaw(): no free id available |
| 18:30 |
erle |
WhiskeyPotato is a mod spawning lots of entities? |
| 18:30 |
WhiskeyPotato |
But I have no idea how to identify what has so many objects to correct it |
| 18:30 |
WhiskeyPotato |
Maybe? |
| 18:30 |
erle |
post mods |
| 18:30 |
WhiskeyPotato |
My server runs over 300 mods |
| 18:30 |
WhiskeyPotato |
and is over 4 years old |
| 18:30 |
erle |
are you getting several megabytes of log per second? |
| 18:31 |
sfan5 |
60000 is a lot of active objects to have so you should notice simply by your fps dropping sharply once you enter the relevant area |
| 18:31 |
WhiskeyPotato |
not quite that much but a lot |
| 18:31 |
erle |
WhiskeyPotato so some mods are stupid and if they can't add something they try it again |
| 18:31 |
WhiskeyPotato |
I could see that |
| 18:32 |
erle |
sfan5 do they need to be active all at once in an area? |
| 18:32 |
sfan5 |
what |
| 18:33 |
WhiskeyPotato |
So im currently in the process of attempting to move from 0.4x to a multicraft base to be able to keep the server up and accessable to the playerbase it has while not having to fix ALL the 300+ mods at once |
| 18:33 |
erle |
well, are the ids immediately recycled? |
| 18:33 |
WhiskeyPotato |
If i could just figure out what is trying to make a new object that would help a lot |
| 18:33 |
erle |
WhiskeyPotato there is probably no easy way to do it. but you can ask your-land.de for their testing strategy. |
| 18:33 |
erle |
they have a shit ton of mods and very little crashes |
| 18:35 |
erle |
WhiskeyPotato you could make a very small mod that hooks into entity creation and logs the parameters |
| 18:35 |
erle |
and maybe which mod is trying to do this |
| 18:35 |
erle |
like, a wrapper function that identifies the culprit |
| 18:36 |
|
est31 joined #minetest |
| 18:36 |
WhiskeyPotato |
im not even sure where to start but i think that might be super helpful. I'll try to unstupid myself for a bit and do that. |
| 18:36 |
sfan5 |
if you have verbose logging enable just search for scriptapi_luaentity_add |
| 18:36 |
erle |
that's probably a more useful suggestion than mine |
| 18:37 |
sfan5 |
if that's infeasible then it's absolutely possible for a mod to go through all current objects and list them to you |
| 18:37 |
WhiskeyPotato |
Dont think it's on verbose at the moment but i'll switch that over and see what we get. of course its an intermittent problem.... |
| 18:37 |
WhiskeyPotato |
I would love to have a mod that does that. maybe i can figure out such. |
| 18:38 |
WhiskeyPotato |
Thanks so much for the suggestions, gonna stew on that for a bit and see what i can come up with. |
| 18:42 |
WhiskeyPotato |
sfan5, can you step me though how a mod could go through all current objects? I'm thinking it would be reading the global table right? so far my luck with reading those has been terrible but my LUA is poo. |
| 18:44 |
erle |
don't worry, lua is a such a quality language that *everyone's* lua is poo |
| 18:50 |
|
YuGiOhJCJ joined #minetest |
| 18:51 |
|
Verticen joined #minetest |
| 19:01 |
sfan5 |
WhiskeyPotato: http://sprunge.us/0ScZdV?lua |
| 19:04 |
WhiskeyPotato |
thank you so much |
| 19:12 |
MTDiscord |
<luatic> erle: No, Lua isn't poo. |
| 19:18 |
erle |
luatic [[triggered]] :P |
| 19:19 |
MTDiscord |
<luatic> yes |
| 19:19 |
|
___nick___ joined #minetest |
| 19:19 |
erle |
sfan5, btw you mentioned a while back that you would make a game. contentdb does not show a game from you. where is it? did you reconsider? |
| 19:19 |
MTDiscord |
<luatic> sfan5 made a game, he just didn't put it on CDB |
| 19:20 |
MTDiscord |
<luatic> check their GH acc |
| 19:20 |
sfan5 |
https://github.com/sfan5/minetest_classic |
| 19:20 |
sfan5 |
putting it on CDB is planned™ |
| 19:20 |
MTDiscord |
<luatic> ninja'd |
| 19:21 |
|
___nick___ joined #minetest |
| 19:21 |
erle |
thx |
| 19:21 |
erle |
finally, oerkki will be back |
| 19:22 |
MTDiscord |
<ROllerozxa> does it feature :dungeonmaster: too? |
| 19:22 |
sfan5 |
all mobs are included |
| 19:23 |
MTDiscord |
<ROllerozxa> awesome |
| 19:23 |
sfan5 |
even ones you can only spawn by cheating yourself the right item (firefly) |
| 19:25 |
erle |
x2048 the thing does not compile, what am i missing? https://mister-muffin.de/p/PV32.txt |
| 19:27 |
erle |
x2048 my mistake, i seem to have bungled it. new error: main.c:5:10: fatal error: GLUT/glut.h: Datei oder Verzeichnis nicht gefunden |
| 19:27 |
erle |
i do have libglu1-mesa-dev |
| 19:28 |
sfan5 |
glu and glut are not the same libraries |
| 19:28 |
erle |
yeah but “apt-file search GLUT/glut.h“ does not give results |
| 19:28 |
erle |
neither does “apt-cache search glut.h“ |
| 19:28 |
erle |
maybe ubuntu has it, but debian does not? |
| 19:29 |
MTDiscord |
<x2048> OK, just remove GLUT/ |
| 19:29 |
MTDiscord |
<x2048> freeglut3 installs glut.h into /usr/include/GL |
| 19:29 |
MTDiscord |
<x2048> or replace GLUT/ with GL/ |
| 19:29 |
MTDiscord |
<x2048> should also work |
| 19:29 |
erle |
hey yeah it works |
| 19:29 |
erle |
lots of warnings during compiling, but it works |
| 19:30 |
MTDiscord |
<x2048> yes, that's what I saw too. |
| 19:30 |
MTDiscord |
<x2048> I guess you want to tweak it? |
| 19:31 |
erle |
#include <math.h> makes some warning go away |
| 19:32 |
MTDiscord |
<x2048> string.h too |
| 19:32 |
erle |
x2048 there is two things: first i want to figure out if i can have shadows using shaders, apparently i can. then i want to figure out some rendering bugs i have been having. do you see a moire pattern on front of the cubes when the light source is behind the cubes? |
| 19:33 |
|
appguru joined #minetest |
| 19:34 |
erle |
x2048 this looks wrong https://mister-muffin.de/p/UBot.png |
| 19:34 |
MTDiscord |
<x2048> moire pattern is there when a face transitions from light to shadow |
| 19:34 |
erle |
so it's an implementation mistake? |
| 19:35 |
erle |
next questions: what qualitatively distinguishes this solution from minetest shadows? i.e. why does this work and minetest shader-based shadows do not? |
| 19:35 |
erle |
on my GPU |
| 19:35 |
erle |
or how can i find out more about it? |
| 19:36 |
erle |
i wonder, could it be that the reason is not shadows at all, but some other shader? |
| 19:36 |
erle |
or something else entirely? |
| 19:36 |
erle |
the amount of geometry? |
| 19:38 |
MTDiscord |
<x2048> Here's how it looks for me: https://imgur.com/j7DP5dI |
| 19:40 |
erle |
direct link to image https://i.imgur.com/j7DP5dI.png |
| 19:41 |
erle |
(imgur.com makes my XSS detector go haywire) |
| 19:45 |
MTDiscord |
<x2048> What I spotted is the use of DEPTH_COMPONENT texture in this example, and we use R32 texture for the shadow map. Your hardware may support the former and not the latter. |
| 19:45 |
MTDiscord |
<x2048> This is for example the case with GLES2 hardware |
| 19:46 |
|
sunyibo left #minetest |
| 19:46 |
erle |
x2048 what would be the visual impact of just using the lower-spec one? |
| 19:48 |
erle |
the hardware is an intel 945GM, which supports OpenGL 1.4 with shaders and OpenGL ES 2.0, GLSL ES 1.0.16 |
| 19:49 |
MTDiscord |
<x2048> Self-shadowing/moire. This is clearly visible if shadow map textures are switched to 16-bit |
| 19:50 |
MTDiscord |
<x2048> (in minetest settings) |
| 19:50 |
erle |
what is the cause of the moire? |
| 19:52 |
erle |
is the depth test not accurate enough? |
| 19:53 |
erle |
x2048 also, i thought that culling the front face (like done in the sm example) would prevent self shadowing |
| 19:56 |
erle |
x2048 in FragmentShader.c there is a hack to reduce moire pattern. it adds 0.0005, but if i remove a zero, then it works. |
| 19:59 |
erle |
it looks gnarly anyway |
| 20:01 |
MTDiscord |
<x2048> Here is how it look in Minetest with 16bit SM: https://mister-muffin.de/p/c3UJ.png |
| 20:02 |
MTDiscord |
<x2048> We do front-face culling with nodes, but not with entities, because it kills proper self-shadowing |
| 20:02 |
MTDiscord |
<x2048> Like shadows inside the plane's cabin |
| 20:02 |
erle |
x2048 looks good enough to me, but the plane has tripes |
| 20:02 |
erle |
strips |
| 20:03 |
erle |
stripes damn |
| 20:04 |
MTDiscord |
<x2048> also notice the sand nodes behind the plane having weird dark spots at the corners |
| 20:05 |
MTDiscord |
<x2048> As for the offsets, here's our code: https://github.com/minetest/minetest/blob/master/client/shaders/nodes_shader/opengl_vertex.glsl#L236-L257 |
| 20:05 |
MTDiscord |
<x2048> it goes well beyond simple z-offset and combines normal offset with Z offset and some heuristic for plant-like nodes without normals |
| 20:06 |
MTDiscord |
<x2048> because we also have that |
| 20:06 |
MTDiscord |
<x2048> and our shadow map is non-uniform, because that gives better detail. |
| 20:11 |
MTDiscord |
<x2048> As for 24 bit depth component, it may be good enough for very-low - low, see 16-bit with very low: https://mister-muffin.de/p/wSFc.png |
| 20:13 |
MTDiscord |
<x2048> The problem is GLES2 does not allow to read the depth buffer, i.e. I cannot bind DEPTH_COMPONENT[24] to a texture, so I'll have to use a 8888 texture and combine the components, which does not work with incremental map updates we are doing today. |
| 20:14 |
erle |
what beyond the moire is the reason for the depth buffer to need that amount of bits? |
| 20:15 |
erle |
i mean, i have no idea of this, i just thought maybe you can blur and mask stuff and get away with it |
| 20:18 |
MTDiscord |
<x2048> with low depth precision entire world will be covered with these artifacts, bleeding light and shadows, moire etc. |
| 20:18 |
MTDiscord |
<x2048> You can blur away everything ? |
| 20:20 |
MTDiscord |
<x2048> can't, I mean |
| 20:28 |
erle |
i just tested my intuitions with the shadow mapping example |
| 20:28 |
erle |
and the world indeed becomes a gray blob if i overdo it |
| 20:29 |
erle |
at lower resolutions the edges of objects bleed shadows out into the void |
| 20:29 |
MinetestBot |
[git] TurkeyMcMac -> minetest/minetest: Let core.get_mod_storage be called multiple times (#12572) d631f21 https://github.com/minetest/minetest/commit/d631f21024bac626babe2e8eac4505210d368ffa (2022-07-23T20:27:07Z) |
| 20:29 |
MinetestBot |
[git] sfan5 -> minetest/minetest: Bump IrrlichtMt 71f6a5f https://github.com/minetest/minetest/commit/71f6a5f44e0cb0e4ef57e942987ec8e76139452c (2022-07-23T20:27:19Z) |
| 20:29 |
MinetestBot |
[git] sfan5 -> minetest/minetest: Centralize IrrlichtMt version used for builds 175e132 https://github.com/minetest/minetest/commit/175e13257653b07df247422cebd3a6da5f29dc51 (2022-07-23T20:27:19Z) |
| 20:29 |
MinetestBot |
[git] sfan5 -> minetest/minetest: Check minimum Irrlicht version before build 7e78c01 https://github.com/minetest/minetest/commit/7e78c0171ea2fec3594656a6abe44695b4252ba1 (2022-07-23T20:27:19Z) |
| 20:53 |
|
YuGiOhJCJ joined #minetest |
| 21:00 |
|
specing_ joined #minetest |
| 21:12 |
|
proller joined #minetest |
| 21:26 |
|
fluxionary joined #minetest |
| 21:27 |
|
natewrench joined #minetest |
| 21:31 |
|
sys4 joined #minetest |
| 21:40 |
Pexin |
0 |
| 21:40 |
muurkha |
-1 |
| 21:43 |
erle |
one, two, three, pi, two pi, apple pie, apple strudel, turd sandwich |
| 21:44 |
|
WhiskeyPotato joined #minetest |
| 21:44 |
|
Verticen joined #minetest |
| 22:13 |
muurkha |
ew |
| 22:21 |
|
vampirefrog joined #minetest |
| 22:25 |
|
dabbill joined #minetest |
| 22:34 |
|
panwolfram joined #minetest |
| 22:45 |
|
natewrench joined #minetest |
| 22:45 |
natewrench |
hey is minetest as a whole thinking about listing themselves on steam/gog like openttd |
| 22:47 |
wsor |
someone would need to cough up the $100 at a minimum |
| 23:02 |
|
dabbill joined #minetest |
| 23:32 |
|
WhiskeyPotato joined #minetest |
